1. Field of the Invention
This invention relates generally to wideband code division multiple access (WCDMA) for a communication system and more particularly to a method of initial frequency acquisition following turn-on for a mobile terminal in a WCDMA system.
2. Description of the Prior Art
Wireless networks are becoming increasingly popular, and in this regard there have been improvements in many aspects of such networks. Some improvements relate to carrier frequency acquisition techniques. F. Classen and H. Meyr, Maximum Likelihood Open Loop Carrier Synchronizer For Digital Radio, IEEE International Conference on Communications (ICC), vol. 1, pp. 493-497 (1993), for example, discloses one method of frequency acquisition implemented for small carrier frequency offsets in which phase differences between consecutive symbols are computed. Classen et al. show that a small carrier frequency offset can be estimated by:             2      ⁢              xe2x80x83            ⁢      π      ⁢              xe2x80x83            ⁢      Δ      ⁢              xe2x80x83            ⁢              f        ^            ⁢      P        =          angle      ⁡              (                              ∑            n                    ⁢                      xe2x80x83                    ⁢                                                    a                                  n                  +                  1                                *                                            a                n                *                                      ⁢                          z              n              *                        ⁢                          z                              n                +                1                                                    )              ;
Zn denotes received symbols, an denotes the data modulating these symbols, P denotes the symbol period and xcex94{circumflex over (f)} an estimate of the carrier frequency offset.
The initial carrier frequency offset seen by a mobile terminal following power-on, can however, be as much as 10 ppm, which is 20 kHz at a carrier frequency of 2 GHz. This large carrier frequency offset is unavoidable where inexpensive oscillators without temperature compensation must be used in order to compete in the wireless network market place. Oscillators having better frequency stability require temperature compensation capability and are prohibitively expensive. In view of the foregoing, there will be as much as a 10 ppm error in the sampling clock at a mobile terminal where the carrier frequency clock and the sampling rate clock are locked together in order to correct the sampling rate offset. A sampling rate of 3.84 MHz will thus see an error as large as 38.4 chips per second. Therefore, the sampling rate will automatically be corrected when the carrier frequency offset is corrected.
The foregoing large errors in the carrier frequency and sampling rate make initial frequency acquisition difficult. The Primary synchronization channel (SCH) associated with WCDMA communication has symbols that are 256 chips (66.7 xcexcsec) long. Further, a carrier frequency offset of 20 kHz for a 2 GHz carrier frequency causes a phase rotation of 480 degrees from the beginning to the end of the PSC symbol. A coherent summing of all 256 chips should not therefore be performed with such a large carrier frequency offset. A sampling rate offset of 38.4 chips per second, for example, causes the sampling position to change by 0.384 chips in 10 msec. Thus, some form of drift compensation would have to be employed to track the correct sampling position if the same sampling position is used for more than 10 msec.
In view of the foregoing, a cost competitive and efficient technique for achieving initial frequency acquisition following power-on for a mobile terminal associated with WCDMA communications is both desirable and necessary to advance the art related to wireless networks.
The present invention is directed to a method of determining a carrier frequency offset immediately following power-up of a mobile terminal during WCDMA mode communication between a base station and the mobile terminal such that the initial carrier frequency can be acquired by the mobile terminal prior to initiating communication. Since the Primary SCH cannot be summed coherently over its full length with a frequency offset as high 20 kHz, it is first divided into parts, i.e. 4 parts, to provide a desired pull-in range for a frequency estimator according to one embodiment of the present invention. In this embodiment, let the correlation outputs with each of the 4 parts of the Primary SCH (each with 64 chips) at position k and time slot m be represented by S1,k,m, S2,k,m, S3,k,m, and S4,k,m. The phase difference xcex94xcfx86 between each segment can then be calculated and used to estimate the carrier frequency offset.                               Δ          ⁢                      xe2x80x83                    ⁢                      f            ^                          =                  Δ          ⁢                      xe2x80x83                    ⁢                      φ            ^                    ⁢                      1                          2              ⁢                              xe2x80x83                            ⁢              π              ⁢                              xe2x80x83                            ⁢              T                                                          (        1        )            
The symbol T denotes the length of a single segment and 1/(2T) is assumed to be 30,000 for one embodiment described herein. The estimate of the phase difference calculated from the Primary SCH""s at position k in one frame is:                               Δ          ⁢                      xe2x80x83                    ⁢                      φ            ^                          =                  angle          ⁡                      (                                                            ∑                                      m                    =                    1                                    L                                ⁢                                  xe2x80x83                                ⁢                                                      s                                          1                      ,                      k                      ,                      m                                        *                                    ⁢                                      s                                          2                      ,                      k                      ,                      m                                                                                  +                                                s                                      2                    ,                    k                    ,                    m                                    *                                ⁢                                  s                                      3                    ,                    k                    ,                    m                                                              +              …              +                                                s                                                            y                      -                      1                                        ,                    k                    ,                    m                                    *                                ⁢                                  s                                      y                    ,                    k                    ,                    m                                                                        )                                              (        2        )            
where L is the number of time slots and y is the number of segments. The position p of the path with the largest magnitude is first estimated by finding                     p        =                                            arg              ⁢                              xe2x80x83                            ⁢              max                        k                    ⁢                      (                                          "LeftBracketingBar"                                                                            ∑                                              m                        =                        1                                            L                                        ⁢                                          xe2x80x83                                        ⁢                                                                  s                                                  1                          ,                          k                          ,                          m                                                *                                            ⁢                                              s                                                  2                          ,                          k                          ,                          m                                                                                                      +                                                            s                                              2                        ,                        k                        ,                        m                                            *                                        ⁢                                          s                                              3                        ,                        k                        ,                        m                                                                              +                  …                  +                                                            s                                                                        y                          -                          1                                                ,                        k                        ,                        m                                            *                                        ⁢                                          s                                              y                        ,                        k                        ,                        m                                                                                            "RightBracketingBar"                            2                        )                                              (        3        )            
The position p of the path with the largest magnitude determined from equation (3) is then used to estimate the phase difference given by equation (2). This phase difference determined by equation (2) is then used in equation (1) to estimate the carrier frequency offset. Finally, the acquisition time is determined by the probability that the path at position p really exists and by the standard deviation of the frequency estimate when the path does exist.
Initial frequency acquisition can also be achieved using a frequency bin method by first correcting the carrier frequency offset to within 2 ppm. This technique is useful where absolutely no additional increase in hardware can be tolerated. The frequency range of interest is first divided into a number of bins, preferably 5 bins, e.g. xe2x88x9216,xe2x88x928, 0, 8 and 16 kHz, which corresponds to xe2x88x928, xe2x88x924, 0, 4 and 8 ppm respectively for a frequency range between xe2x88x9220 kHz and +20 kHz. Frequency acquisition is performed by first assuming the carrier frequency offset is 0 Hz and performing the steady state acquisition. If the carrier frequency offset is outside the range of xe2x88x924 to 4 kHz, then the acquisition will likely fail, and another bin will be tried. The mobile terminal preferably stores carrier frequency offset information from just before it was previously powered-off. During power-on, the mobile terminal can then start the search in the bin corresponding to the one stored in memory.
Accordingly, a WCDMA communication signal is first downconverted and sampled at the receiver or mobile terminal. Frequency acquisition is then implemented in 3 stages. In stage 1 of acquisition, the primary synchronization code is first located. During stage 2 of acquisition, the secondary synchronization code that overlaps the primary synchronization channel containing the primary synchronization channel symbols is decoded. This secondary synchronization code indicates the code group used by the cell. In stage 3 of acquisition, the particular scrambling code used by the base station of interest is determined by searching through the scrambling codes in the code group. If the correct scrambling code has been determined, the receiver can enter its steady state operating mode wherein communication between the base station and the mobile terminal takes place. If communication between the base station and the mobile terminal cannot be achieved within a predetermined period of time, another bin is chosen and the foregoing process repeated. This process will then continue until the local VCO frequency is correctly adjusted to account for the carrier frequency offset thereby allowing the mobile terminal/receiver to enter its steady state operating mode.
As used herein, the following words have the following meanings. The words xe2x80x9calgorithmic softwarexe2x80x9d means an algorithmic program used to direct the processing of data by a computer or data processing device. The words xe2x80x9cdata processing devicexe2x80x9d as used herein refer to a CPU, DSP, microprocessor, micro-controller, or other like device and an interface system. The interface system provides access to the data processing device such that data could be entered and processed by the data processing device. The words xe2x80x9cdiscrete dataxe2x80x9d as used herein are interchangeable with xe2x80x9cdigitized dataxe2x80x9d and xe2x80x9cdigitized dataxe2x80x9d as used herein means data which are stored in the form of singularly isolated, discontinuous data or digits.
In one aspect of the present invention, a method of achieving initial frequency acquisition for a mobile terminal in a WCDMA system accommodates frequency offsets up to about 10 ppm.
In another aspect of the present invention, a method of achieving initial frequency acquisition for a mobile terminal in a WCDMA system accommodates frequency offsets up to about 10 ppm with no additional hardware requirements.
In still another aspect of the present invention, a method of achieving initial frequency acquisition for a mobile terminal in a WCDMA system accommodates frequency offsets up to about 10 ppm with substantially no user noticeable operational characteristics.
In yet another aspect of the present invention, a method of achieving initial frequency acquisition for a mobile terminal in a WCDMA system without requiring use of a temperature compensated oscillator at the mobile terminal.
Still another aspect of the present invention is associated with achieving initial frequency acquisition for a mobile terminal in a WCDMA system that is easy to accomplish and cost competitive in the wireless network product marketplace, even when the carrier frequency offset between a base station and the mobile terminal is as large as about 10 ppm.
Yet another aspect of the present invention is associated with achieving initial frequency acquisition for a mobile terminal in a WCDMA communication system without use of multiple metrics such that memory usage efficiency is optimized.